Prompt Engineering: New Career Path or Flash in the Pan?  

Introduction 

In the two years or so since its launch, there’s a very good chance you’ve had a conversation with ChatGPT at least once. 

But did you know that, as you keyed in those instructions, you were unwittingly training yourself in the art of the “prompt engineer”?  

With so much buzz around generative AI, there are questions about whether or not the concept of prompt engineering is a genuine career opportunity or just a fleeting distraction during the AI gold-rush.  

Here’s what you need to know to make up your own mind.  

What is prompt engineering? 

Prompt engineering is the art (or the science) of writing text-based instructions, known as ‘prompts’, to produce the best possible result from a generative AI tool.  

This discipline is about crafting instructions which get AI models, like OpenAI’s GPT 4, to produce spot-on responses. When it’s done right, prompt engineering can transform your AI interactions from meh to marvellous.  

A well-crafted prompt is like a magic wand that gets the AI to understand exactly what you want, capturing the context, tone, and nuances perfectly. It can save time, resources, and become an essential skill for any business.  

What does a prompt engineer actually do? 

So, what does a prompt engineer actually get up to all day? 

It’s much more than simply throwing questions at a Large Language Model (LLM) and hoping for the best. The role of a prompt engineer is to understand how these AI models think, figuring out their quirks and capabilities then writing prompts and instructions which will generate the best possible output.  

A prompt engineer will write, tweak, and test prompts until they perfect the formula. It’s a blend of creativity, technical know-how, and a hefty dose of patience. They also keep an eye on the latest AI trends and work together with developers and data scientists to uncover ways to weave AI magic into bigger projects. 

Here are a few examples of the day-to-day tasks a prompt engineer might deal with: 

• Writing and testing custom instruction sets. This involves creating specific instructions tailored to particular tasks or industries. The prompt engineer tests these instructions rigorously to ensure the AI delivers accurate and relevant responses. 
• Refining prompt templates for consistency. Standardising prompts across different applications helps maintain a consistent AI performance. This task requires the prompt engineer to continuously refine and update templates based on feedback and performance data. 
• Collaborating on AI training projects. Working with data scientists and developers, prompt engineers contribute to training AI models by providing expertly crafted prompts. This collaboration is crucial for developing robust AI systems that understand and respond consistently to different inputs. 
• Analysing AI output for improvements. Prompt engineers regularly review the AI’s responses to identify areas for enhancement. By understanding where the AI might be falling short, they can adjust prompts to improve future responses. 
   

What skills does a prompt engineer need? 

Because the role is so new, there are no formal qualifications needed to become a prompt engineer. 

That said, there are some traits common to the most successful AI prompt writers. 

• First off, linguistic proficiency is key. A strong command of language and an understanding of semantics are crucial for crafting clear and effective prompts.  
• Next, you’ll need some technical know-how. Familiarity with AI and machine learning concepts will help you grasp how generative models work and allow you to tailor your prompts to the AI’s strengths… and limitations.  
• A problem-solving mindset is essential, too. Being able to analyse the AI’s responses and identify areas for improvement means you can continually refine your prompts for better results.  
• Finally, creativity is a must. Crafting prompts that push the boundaries of what AI can produce requires innovative thinking and a willingness to try a new approach just to see what happens. 
   

Do I need to understand prompt engineering to get results with AI?  

So, do you really need to dive into the nitty-gritty of prompt engineering to make the most out of AI? The short answer is… probably not. Thanks to tools like Microsoft Copilot, AI is becoming more user-friendly than ever.  

These tools come with canned prompts and aim to make AI as simple as possible, often offering one-click solutions rather than requiring users to craft detailed prompts. For example, Copilot in Word allows you to highlight a section of text, then have the AI rewrite it at the click of a button. This means you can still get excellent results from AI without having to write a single prompt. 

That said, a basic understanding of prompt engineering can certainly give you an edge. Knowing how to tweak and refine prompts allows you to get even more tailored and precise outputs from the AI. It’s a bit like having a cheat code-sure, you can play the game without it, but having it makes things a whole lot smoother and more fun. 

Is prompt engineering here to stay? 

Now, for the big question: is prompt engineering just a flash in the pan or can it become a real career with real staying power?  

At Get Support, we believe that, as AI continues to evolve, so too will the need for savvy prompt engineers. The ‘engineer’ moniker might not stick, but the skills involved almost certainly will. We might even see these skills become essential across various fields, from healthcare to finance, where knowing the industry-specific jargon and helping AI make sense of it can makes all the difference.  

It’s still early days for generative AI, but the landscape is wide open, and the demand for these skills is only set to grow.  

If you haven’t yet dipped your toe into the world of generative AI at work, why not ask your Get Support account manager about what’s possible with Microsoft Copilot and other tools?